This proposal has as its objective the development of more selective therapy for hepatic tumors and high grade glimas by exploiting and combining two therapeutic strategies, namely, regional delivery of chemotherapy and biochemical modulation of antimetabolite action. We have selected for study a group of antimetabolites whose pharmacokinetics properties should permit the maintenance of a substanially higher concentration at the site of the tumor in the general circulation and, whose biochemical mechanisms invite potentially selective modulation of cytotoxicity by co-administration with other agents. These drugs are: 5-bromo-2'-deoxyuridine (BrdUrd), 5-iodo-2'-deoxyuridine (IdUrd), 6-thioguanine (TG) and 6-mercaptopurine (MP) - all nucleic acid precursor analogs where a crucial target effect is incorporation into DNA. Monitoring this target parameter with recently developed sensitive analytical techniques should make it feasible to correlate effects in cultured cells to those in relevant animal and human tissue in vivo. Variables to be studied across the spectrum from cell culture to the clinic as affecting the selectivity of target incorporation and cytotoxicity include (for each agent): influence of concentration, duration of exposure, and effects of modulating agents used to influence levels of relevant cellular metabolites. It is felt that examination of human tumor type specific cultured cells (colon-Project 1A, gliomas-Project 1B) may reveal common themes as well as define, to a degree, the heterogeneity relevant to these pertinent cancers in vivo. The potential ability to extrapolate from an in vitro cell culture to an in vivo setting will be explored in the VX2 rabbit tumor model (Project 2). This model will be used to define the selective advantage of regional infusions in achieving higher regional and lower systemic exposures as well as the ability of antimetabolite modulation to selectively change analog DNA incorporation in tumor versus normal host tissues. Resultant antitumor effect and host toxicity will be assessed relative to the selectivity of measured analog incorporation into tissue DNA. Project 3 evaluates the pharmacokinetics and toxicity of regional administration as well as of antimetabolite modulation as a means to and interesting in vitro regimens into rationally designed clinical studies. Phase 1/clinical pharmacology studies (Project 4) involving patients with hepatic cancers examine hepatic arterial chemotherapy and antimetabolite modulation as a means to extend regimens of potentially improved activity/selectivity from the preclinical models into the clinic in a relevant, controlled fashion. Project 5 similarly extends regimens from preclinical models into phase I-II studies in patients with high grade gliomas, examining the interaction of antimetabolite treatments with radiotherapy (both external beam and interstitial). It is anticipated that the approach taken in this proposal will lead to productive bidirectional interactions between preclinical and clinical investigation, validating at least some of the methods chosen and leading to more creative, rational, and potent clinical therapies utilizing these antimetabolites.